This updates my August 6, 2016, report of Republican underperformance, relative to Democrats, in absentee voting in the 8/30/2016 Broward primary election. I showed that, by controlling for earliness and geography, the relative underperformance was a partisan trend not explained by obvious artifacts. The cumulative trend continues but has reduced (see graph).
In the most recent data, 8/15/2016, 15 days before this 8/30/2016 primary election, absentee votes as a ratio of all uncancelled absentee requests were Democrats, 29.9%; Republicans, 22.8%; and Independents (other parties and no party affiliation), 16.8%. In the 8/14/2012 primary, the ratios 15 days before the election were Dems, 21.3%; Reps, 18.9%; and Indies, 12.8%.
In the 2016 primary, the deficit in Republican voting performance, relative to Democrats, narrowed in the past 10 days from 41% to 24%. This shows a bounce back for Republicans, which will still likely not erase the cumulative underperformance by 8/30/2016.
The same data can be used to argue that there is not Republican underperformance but Democratic overperformance. Total Democratic votes to date were 23,546 compared with 11,235 on the 15th day before the election in 2012. The absentee votes by Republicans 15 days before the 2016 and 2012 primaries were 10,077 and 6,081, respectively.
Uniformly extending the last 7 days of absentee voting through 8/30/2016 projects a total absentee vote for Dems, 42,877; Reps, 20,919; and Indies, 7,805 compared with 2012 which was Dems, 24,472; Reps, 13,580; and Indies 4,064.
This is no longer as flaming good news for Democrats as it was, but, without a doubt, Democrats have learned very well how to vote absentee, something that the Republicans used to clobber us with. Just as Republicans learned early voting in the 2010 primary, in 2014 and 2016 Democrats really learned absentee voting.